demonstration of the analytical performance

DEMONSTRATION OF THE ANALYTICAL PERFORMANCE OF THE NAVIOS FLOW CYTOMETER IN A MULTI-CENTER STUDY
Diana Careaga1, Robert Magari1, Karen Lo1, Michael Keeney2, Janice Popma2, Joanne Luider3, Aito Ueno3, Gerard Lozanski4, Bruce Briggs4, Elena Afonina5, Enrique Rabellino5, and Liliana Tejidor1
Beckman Coulter Diagnostics, Inc., 1Clinical Research, Miami, Florida, USA
2London Health Sciences Centre, London, Ontario, Canada
3Foothills Hospital, Calgary, Alberta, Canada
4Ohio State University Medical Center, Columbus, Ohio
Beckman Coulter Life Science, Inc., 5Clinical Applications Development, Miami, Florida, USA
Navios Tetra Precision Performance using control materials
- tetraCHROME CD45-FITC/CD4-RD1/CD8-ECD/CD3-PC5 Reagent
Introduction
CYTO-STATR
The performance of the Navios™ cytometer* with
tetraCHROME™ reagents compared to the Cytomics FC500 for
lymphocyte subset analysis in patients suspected of immunodeficiency
and apparently healthy individuals was evaluated in a multi-center study.
Objectives and Methods
To assess accuracy, the Navios Tetra method was compared to both
FC500 manual gating and the tetraCXP algorithm methods in ≥300
samples covering the CD4+ medical decision levels at 3 sites. Samples
were prepared by using manual pipetting and TQ-Prep or Prep Plus 2 and
TQ-Prep.
Precision performance of the Navios CYTO-STATR tetraCHROME™
system was evaluated based on CLSI EP5-A2 using control materials.
Combined Sites
ImmunoTrol
Analyte
Unit
N
cells/µL
244
Total CD3+
%
244
cells/µL
244
CD3+CD4+
%
244
cells/µL
244
CD3+CD8+
244
%
ImmunoTrol
Analyte
Unit
cells/µL
Total CD3+
%
cells/µL
CD3+CD4+
%
cells/µL
CD3+CD8+
%
Low
N
245
245
245
245
245
245
Mean
746
72.32
483
46.80
247
23.98
PrepPlus 2 and TQ-Prep
Repeatability Within Site Reproducibility
%CV
%CV
%CV
3.696
5.106
5.360
0.855
0.988
1.003
3.890
5.209
5.461
1.520
1.606
1.655
4.569
5.831
6.092
2.920
2.920
2.939
Mean
344
57.82
133
22.36
194
32.66
Repeatability Within Site Reproducibility
%CV
%CV
%CV
3.355
4.505
5.122
1.632
1.661
1.669
4.422
5.712
6.858
3.045
3.345
3.526
3.789
4.756
5.519
2.545
2.556
2.565
Table 1. Precision results from the combined sites
FC 500 Manual (Reference – FC Man) vs. Navios Tetra
(Test – Navios) and FC500 tetraCXP (Test – FC)
Figure 1. CD3+CD4+ Absolute Count – Estimated Bias & Confidence Intervals Plot
The adult reference interval was determined following CLSI EP28-A2.
Results
The Navios™ cytometer showed excellent repeatability & reproducibility.
Table 1 summarizes representative results for tetraCHROME tube 1 for
all sites combined. The repeatability and reproducibility for very low CD4
levels (133 cells/µL) was <5% and <7%, respectively.
Figure 1 illustrates a three-way comparison (Navios vs. FC500 manual
and FC500 CXP vs. FC500 manual) confirming the improved
comparability of Navios tetra to the manual reference method (FC500).
The comparison
of Navios Tetra to FC500 manual gating showed
minimal positive bias at medical decision levels and slightly higher
positive bias when compared to tetraCXP.
Bias and confidence intervals for these comparisons are shown in Tables
2 & 3 while Bland Altman plots are illustrated in Figures 2 & 3. Observed
differences are in part the result of modifications made to the Navios tetra
algorithm compared to the FC500
algorithm (tetraCXP) to include
lymphocyte populations with altered forward light scatter properties that
maintain both side scatter and CD45 fluorescence intensity properties.
The inclusion of these populations is consistent with published literature
(Glencross1 et. al., ).
The analysis method employed in Navios tetra presumes that the sample
at time of collection contains minimal apoptosis-committed cells since
apoptosis progresses quickly and apoptotic cells are cleared from the
circulation very rapidly. Any apoptosis and eventual cell death in the
specimen is more likely the result of pre-analytical conditions postvenipuncture such as temperature and the effects due to aging of the
specimens. Some samples demonstrating decreased forward scatter
properties that may be characteristic of dead or apoptotic cells may yield
different results from previous tetra algorithms.
Navios Tetra (Test) vs. FC 500 Manual (Reference)
- tetraCHROME CD45-FITC/CD4-RD1/CD8-ECD/CD3-PC5 Reagent
Lower
Upper
Confidence Confidence
Limit
Limit
Conducted at 3 sites in North America
Level
Bias
25th Percentile
1050
-3.34
-8.83
2.14
50th Percentile
1345
-10.21
-17.07
-3.35
75th Percentile
1725
-19.03
-29.19
-8.86
cells/µL
Total CD3+
th
25 Percentile
%
th
0.53
82.12
0.47
0.32
0.62
306
9.25
6.60
11.90
531
8.07
3.35
12.78
75 Percentile
Medical Decision
Medical Decision
Medical Decision
Medical Decision
827
50
100
350
500
6.51
10.59
10.33
9.02
8.23
-1.12
3.01
3.80
6.12
3.77
14.15
18.18
16.86
11.91
12.68
25th Percentile
21.96
0.83
0.68
0.98
50 Percentile
30.84
0.79
0.66
0.91
75th Percentile
42.31
0.73
0.58
0.89
25th Percentile
470
-4.74
-7.84
-1.64
50th Percentile
695
-6.61
-11.09
-2.13
75th Percentile
988
-9.04
-16.30
-1.78
25th Percentile
29.64
-0.01
-0.20
0.17
50th Percentile
42.37
0.11
-0.04
0.25
75th Percentile
56.02
0.24
0.06
0.42
25 Percentile
th
50 Percentile
th
cells/µL
CD3+CD8+
%
th
Table 2. Bias and Its Confidence Intervals at Medical Decision Points and Quartiles
Navios Tetra (Test) vs. FC 500 Manual (Reference)
Level
Bias
25th Percentile
1001
35.12
28.90
41.33
50th Percentile
1294
36.91
29.11
44.72
75th Percentile
1675
39.26
27.58
50.94
25th Percentile
72.93
-0.85
-1.01
-0.69
50th Percentile
78.61
-0.84
-0.98
-0.70
75th Percentile
83.49
-0.82
-0.99
-0.66
25th Percentile
284
23.68
20.88
26.47
50th Percentile
507
33.72
28.60
38.84
75th Percentile
806
47.26
38.76
55.77
Medical Decision
50
13.10
4.50
21.69
Medical Decision
100
15.36
8.10
22.61
Medical Decision
350
26.65
23.50
29.79
Medical Decision
500
33.42
27.52
39.33
25th Percentile
20.71
0.87
0.70
1.03
50th Percentile
30.60
0.60
0.47
0.74
75th Percentile
42.15
0.30
0.14
0.46
25th Percentile
460
12.57
9.40
15.75
50th Percentile
682
16.87
12.25
21.48
75th Percentile
973
22.51
14.98
30.05
25th Percentile
30.62
-0.48
-0.65
-0.31
50th Percentile
43.38
-0.65
-0.78
-0.51
75th Percentile
56.95
-0.83
-1.00
-0.66
cells/µL
Total CD3+
%
cells/µL
CD3+CD4+
%
cells/µL
CD3+CD8+
%
Table 3. Bias and Its Confidence Intervals at Medical Decision Points and Quartiles
Navios Tetra (Test) vs. FC 500 tetraCXP (Reference)
Normal reference intervals for lymphocyte subsets (data not shown) were
consistent with published values.
Conclusion
The Navios™ flow cytometer demonstrated excellent performance in
lymphocyte subset analysis with CYTO-STATR tetraCHROME™.
*Recently received clearance by the US FDA on Sept 24, 2013; available
for in vitro diagnostic use within the U.S.A.
1Glencross,
D.K. et. al. Large-scale Affordable Panleucogated CD4+
testing with Proactive Internal & External Quality Assessment: In Support
of the South African National Comprehensive Care, treatment &
Management Programme for HIV & AIDS
Figure 2. CD3+CD4+ Absolute Count – Bland-Altman Plot
Lower
Upper
Confidence Confidence
Limit
Limit
Conducted at 3 sites in North America
0.47
0.28
th
%
0.18
0.40
75 Percentile
CD3+CD4+
0.33
77.44
50 Percentile
th
cells/µL
71.84
Navios Tetra (Test) vs. FC 500 tetraCXP (Reference)
- tetraCHROME CD45-FITC/CD4-RD1/CD8-ECD/CD3-PC5 Reagent
Figure 3. CD3+CD4+ Absolute Count – Bland-Altman Plot